| { | |
| "best_metric": 3.4680771827697754, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 2.962962962962963, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.014814814814814815, | |
| "grad_norm": 68.60965728759766, | |
| "learning_rate": 1e-05, | |
| "loss": 15.6754, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.014814814814814815, | |
| "eval_loss": 4.450826644897461, | |
| "eval_runtime": 1.3189, | |
| "eval_samples_per_second": 86.436, | |
| "eval_steps_per_second": 21.988, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02962962962962963, | |
| "grad_norm": 61.886592864990234, | |
| "learning_rate": 2e-05, | |
| "loss": 16.7019, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.044444444444444446, | |
| "grad_norm": 76.6452865600586, | |
| "learning_rate": 3e-05, | |
| "loss": 16.8666, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05925925925925926, | |
| "grad_norm": 68.32090759277344, | |
| "learning_rate": 4e-05, | |
| "loss": 16.9719, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 64.47012329101562, | |
| "learning_rate": 5e-05, | |
| "loss": 17.1615, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 59.52012252807617, | |
| "learning_rate": 6e-05, | |
| "loss": 17.0237, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.1037037037037037, | |
| "grad_norm": 64.67552185058594, | |
| "learning_rate": 7e-05, | |
| "loss": 16.8811, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 63.22369384765625, | |
| "learning_rate": 8e-05, | |
| "loss": 16.6728, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.13333333333333333, | |
| "grad_norm": 62.07340621948242, | |
| "learning_rate": 9e-05, | |
| "loss": 17.1658, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 61.928707122802734, | |
| "learning_rate": 0.0001, | |
| "loss": 16.7611, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16296296296296298, | |
| "grad_norm": 56.120635986328125, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 17.198, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 58.08959197998047, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 16.9448, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1925925925925926, | |
| "grad_norm": 57.3941535949707, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 16.8164, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2074074074074074, | |
| "grad_norm": 57.866004943847656, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 17.6413, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 66.68106079101562, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 17.7946, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 80.18011474609375, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 18.7777, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2518518518518518, | |
| "grad_norm": 53.23362350463867, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 15.0291, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 43.26505661010742, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 15.5285, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2814814814814815, | |
| "grad_norm": 45.03691101074219, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 15.6638, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 44.33474349975586, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 15.774, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3111111111111111, | |
| "grad_norm": 47.4934196472168, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 14.8594, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.32592592592592595, | |
| "grad_norm": 51.91562271118164, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 15.6994, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.34074074074074073, | |
| "grad_norm": 63.69327163696289, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 15.1105, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 70.22563171386719, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 15.5335, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 51.048397064208984, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 15.695, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3851851851851852, | |
| "grad_norm": 60.33977508544922, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 15.4985, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 54.047393798828125, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 15.9076, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4148148148148148, | |
| "grad_norm": 52.88368606567383, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 16.171, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.42962962962962964, | |
| "grad_norm": 54.394630432128906, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 16.1892, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 64.61843872070312, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 16.4766, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.45925925925925926, | |
| "grad_norm": 52.421852111816406, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 16.5611, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 74.07872772216797, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 17.0349, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4888888888888889, | |
| "grad_norm": 45.0691032409668, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 14.59, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5037037037037037, | |
| "grad_norm": 43.47343063354492, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 15.0874, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5185185185185185, | |
| "grad_norm": 51.036460876464844, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 14.6667, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 44.959651947021484, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 15.2569, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5481481481481482, | |
| "grad_norm": 53.07027816772461, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 14.7695, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.562962962962963, | |
| "grad_norm": 66.9604263305664, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 15.0859, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.5777777777777777, | |
| "grad_norm": 73.07352447509766, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 15.3251, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 83.22005462646484, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 14.9193, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6074074074074074, | |
| "grad_norm": 95.94990539550781, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 14.8468, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 78.98324584960938, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 15.2293, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6370370370370371, | |
| "grad_norm": 90.03215789794922, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 14.8712, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6518518518518519, | |
| "grad_norm": 108.2412109375, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 15.2733, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 109.66105651855469, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 14.7888, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.6814814814814815, | |
| "grad_norm": 101.620361328125, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 14.7101, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6962962962962963, | |
| "grad_norm": 118.87198638916016, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 15.6875, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 125.9925765991211, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 15.8599, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.725925925925926, | |
| "grad_norm": 91.35040283203125, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 14.0139, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 98.83194732666016, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 14.9001, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "eval_loss": 3.8177599906921387, | |
| "eval_runtime": 1.308, | |
| "eval_samples_per_second": 87.154, | |
| "eval_steps_per_second": 22.171, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7555555555555555, | |
| "grad_norm": 132.87249755859375, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 15.1441, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.7703703703703704, | |
| "grad_norm": 110.09635925292969, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 15.0139, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.7851851851851852, | |
| "grad_norm": 117.48370361328125, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 15.3807, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 106.9300308227539, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 14.9309, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8148148148148148, | |
| "grad_norm": 132.26931762695312, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 14.7938, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 83.11085510253906, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 14.6302, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.8444444444444444, | |
| "grad_norm": 86.0311279296875, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 14.7895, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.8592592592592593, | |
| "grad_norm": 86.85681915283203, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 14.9282, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.8740740740740741, | |
| "grad_norm": 88.0314712524414, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 15.1812, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 83.65817260742188, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 14.9758, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9037037037037037, | |
| "grad_norm": 81.15483856201172, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 14.4124, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9185185185185185, | |
| "grad_norm": 82.79302215576172, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 14.7479, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.9333333333333333, | |
| "grad_norm": 76.19058990478516, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 14.5411, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 93.92029571533203, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 14.5308, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.9629629629629629, | |
| "grad_norm": 80.75183868408203, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 14.6427, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "grad_norm": 93.4598388671875, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 14.5004, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.9925925925925926, | |
| "grad_norm": 97.47583770751953, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 14.9868, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0074074074074073, | |
| "grad_norm": 107.98495483398438, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 14.8387, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0222222222222221, | |
| "grad_norm": 112.4440689086914, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 14.1241, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.037037037037037, | |
| "grad_norm": 93.2449722290039, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 14.3574, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.0518518518518518, | |
| "grad_norm": 97.04306030273438, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 14.2975, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 110.81587219238281, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 14.6306, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.0814814814814815, | |
| "grad_norm": 107.47632598876953, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 13.8618, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.0962962962962963, | |
| "grad_norm": 107.0185546875, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 14.6864, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 100.5071792602539, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 14.0987, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.125925925925926, | |
| "grad_norm": 89.75070190429688, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 15.0281, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.1407407407407408, | |
| "grad_norm": 109.45982360839844, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 14.6123, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.1555555555555554, | |
| "grad_norm": 102.81034088134766, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 14.745, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.1703703703703703, | |
| "grad_norm": 87.77413177490234, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 14.6252, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 116.85652160644531, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 14.4662, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 109.6570053100586, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 14.4813, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.2148148148148148, | |
| "grad_norm": 88.9306869506836, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 15.1554, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.2296296296296296, | |
| "grad_norm": 85.2114486694336, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 15.1204, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.2444444444444445, | |
| "grad_norm": 70.78118133544922, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 14.4312, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.2592592592592593, | |
| "grad_norm": 70.97513580322266, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 14.021, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.2740740740740741, | |
| "grad_norm": 87.23574829101562, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 14.492, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.2888888888888888, | |
| "grad_norm": 90.26329040527344, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 14.6331, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 76.13572692871094, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 15.0053, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.3185185185185184, | |
| "grad_norm": 87.63103485107422, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 14.487, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 64.3780288696289, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 14.5629, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.348148148148148, | |
| "grad_norm": 68.84786987304688, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 14.1251, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.362962962962963, | |
| "grad_norm": 68.6078109741211, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 13.8759, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.3777777777777778, | |
| "grad_norm": 57.65863800048828, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 14.5024, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.3925925925925926, | |
| "grad_norm": 58.59613800048828, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 14.1191, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.4074074074074074, | |
| "grad_norm": 62.90269470214844, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 14.738, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 62.02676773071289, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 13.9563, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.4370370370370371, | |
| "grad_norm": 61.09934616088867, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 14.026, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.4518518518518517, | |
| "grad_norm": 89.50511169433594, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 14.5478, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.4666666666666668, | |
| "grad_norm": 79.56632995605469, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 14.9694, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.4814814814814814, | |
| "grad_norm": 90.40596008300781, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 14.1793, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.4814814814814814, | |
| "eval_loss": 3.61611008644104, | |
| "eval_runtime": 1.3068, | |
| "eval_samples_per_second": 87.235, | |
| "eval_steps_per_second": 22.191, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.4962962962962962, | |
| "grad_norm": 50.52888107299805, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 14.3801, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.511111111111111, | |
| "grad_norm": 75.90869140625, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 13.9113, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.525925925925926, | |
| "grad_norm": 53.31604766845703, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 13.6829, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 52.32286071777344, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 13.531, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 51.600547790527344, | |
| "learning_rate": 5e-05, | |
| "loss": 13.5156, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.5703703703703704, | |
| "grad_norm": 58.452152252197266, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 13.637, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.585185185185185, | |
| "grad_norm": 73.30082702636719, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 13.6241, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 62.18711853027344, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 13.7785, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.6148148148148147, | |
| "grad_norm": 52.56895065307617, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 14.1475, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.6296296296296298, | |
| "grad_norm": 58.66401290893555, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 14.1419, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.6444444444444444, | |
| "grad_norm": 57.37771224975586, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 14.3208, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 53.625038146972656, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 14.3515, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.674074074074074, | |
| "grad_norm": 85.90170288085938, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 13.572, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.6888888888888889, | |
| "grad_norm": 51.710689544677734, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 13.4417, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.7037037037037037, | |
| "grad_norm": 63.94369125366211, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 13.3869, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.7185185185185186, | |
| "grad_norm": 49.40534591674805, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 13.3057, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.7333333333333334, | |
| "grad_norm": 45.39258575439453, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 13.9662, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.748148148148148, | |
| "grad_norm": 55.47205352783203, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 14.3071, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.762962962962963, | |
| "grad_norm": 55.345340728759766, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 13.8117, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 55.53709030151367, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 14.4789, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.7925925925925927, | |
| "grad_norm": 51.58026123046875, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 13.6507, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.8074074074074074, | |
| "grad_norm": 59.68843078613281, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 14.7606, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.8222222222222222, | |
| "grad_norm": 51.65839385986328, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 13.7926, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.837037037037037, | |
| "grad_norm": 48.51106643676758, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 13.3931, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.8518518518518519, | |
| "grad_norm": 54.19108581542969, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 14.2604, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.8666666666666667, | |
| "grad_norm": 69.57347869873047, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 13.8229, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.8814814814814815, | |
| "grad_norm": 53.762916564941406, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 14.213, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 52.337310791015625, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 12.8749, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.911111111111111, | |
| "grad_norm": 55.00204849243164, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 14.0695, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.925925925925926, | |
| "grad_norm": 72.65208435058594, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 13.5056, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.9407407407407407, | |
| "grad_norm": 75.98030090332031, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 13.6172, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.9555555555555557, | |
| "grad_norm": 67.42346954345703, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 13.8268, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.9703703703703703, | |
| "grad_norm": 55.366458892822266, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 13.5119, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.9851851851851852, | |
| "grad_norm": 54.883052825927734, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 13.7602, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 182.34469604492188, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 12.6965, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 55.028873443603516, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 13.2022, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.0296296296296297, | |
| "grad_norm": 51.80314254760742, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 13.439, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.0444444444444443, | |
| "grad_norm": 52.953208923339844, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 13.7329, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.0592592592592593, | |
| "grad_norm": 43.84860610961914, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 13.9878, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.074074074074074, | |
| "grad_norm": 49.66146469116211, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 13.5835, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.088888888888889, | |
| "grad_norm": 50.06467056274414, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 13.3342, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.1037037037037036, | |
| "grad_norm": 57.57233810424805, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 13.3379, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.1185185185185187, | |
| "grad_norm": 50.848915100097656, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 13.4943, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 49.15460968017578, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 13.4682, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.148148148148148, | |
| "grad_norm": 49.58591079711914, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 13.8458, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.162962962962963, | |
| "grad_norm": 41.95174789428711, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 12.8936, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.1777777777777776, | |
| "grad_norm": 44.95238494873047, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 13.3646, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.1925925925925926, | |
| "grad_norm": 55.146297454833984, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 13.8854, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.2074074074074073, | |
| "grad_norm": 45.33646011352539, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 13.5536, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 47.633079528808594, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 14.2463, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "eval_loss": 3.5002286434173584, | |
| "eval_runtime": 1.3075, | |
| "eval_samples_per_second": 87.188, | |
| "eval_steps_per_second": 22.179, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.237037037037037, | |
| "grad_norm": 62.15080261230469, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 14.169, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 43.1384391784668, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 13.35, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.2666666666666666, | |
| "grad_norm": 61.10799789428711, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 13.4743, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.2814814814814817, | |
| "grad_norm": 40.51551818847656, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 12.9637, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.2962962962962963, | |
| "grad_norm": 47.04324722290039, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 13.9894, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.311111111111111, | |
| "grad_norm": 47.283870697021484, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 13.3073, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.325925925925926, | |
| "grad_norm": 46.01675796508789, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 13.5307, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.3407407407407406, | |
| "grad_norm": 46.414791107177734, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 13.6898, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.3555555555555556, | |
| "grad_norm": 49.98985290527344, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 13.7374, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 50.01529312133789, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 12.7889, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.3851851851851853, | |
| "grad_norm": 52.6187858581543, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 13.1989, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 53.43068313598633, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 13.0332, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.414814814814815, | |
| "grad_norm": 46.10429382324219, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 13.9437, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.4296296296296296, | |
| "grad_norm": 48.99384689331055, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 13.1216, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 54.56727600097656, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 13.7897, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.4592592592592593, | |
| "grad_norm": 71.74275207519531, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 14.0059, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.474074074074074, | |
| "grad_norm": 62.40850067138672, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 13.1038, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 51.76642990112305, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 12.9231, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.5037037037037035, | |
| "grad_norm": 46.97688293457031, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 13.44, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.5185185185185186, | |
| "grad_norm": 44.13114547729492, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 13.1346, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.533333333333333, | |
| "grad_norm": 43.373104095458984, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 13.7069, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.5481481481481483, | |
| "grad_norm": 57.0178337097168, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 13.2752, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.562962962962963, | |
| "grad_norm": 47.08359146118164, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 13.388, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.5777777777777775, | |
| "grad_norm": 52.63562774658203, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 13.3634, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.5925925925925926, | |
| "grad_norm": 49.64703369140625, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 13.0749, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 54.43859100341797, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 13.4577, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.6222222222222222, | |
| "grad_norm": 50.75735855102539, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 13.0154, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.637037037037037, | |
| "grad_norm": 51.970638275146484, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 14.282, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.651851851851852, | |
| "grad_norm": 49.63145065307617, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 13.1712, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 48.46242904663086, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 12.8918, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.6814814814814816, | |
| "grad_norm": 56.70124816894531, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 13.4566, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.696296296296296, | |
| "grad_norm": 44.90093994140625, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 12.7927, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.7111111111111112, | |
| "grad_norm": 74.2798843383789, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 12.6853, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 45.66691970825195, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 12.6442, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.7407407407407405, | |
| "grad_norm": 43.10373306274414, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 13.6349, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.7555555555555555, | |
| "grad_norm": 48.92824935913086, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 13.7068, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.7703703703703706, | |
| "grad_norm": 50.793540954589844, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 13.1044, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.785185185185185, | |
| "grad_norm": 49.439361572265625, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 13.2471, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 48.739723205566406, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 12.8476, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.814814814814815, | |
| "grad_norm": 52.2317008972168, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 13.5524, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.8296296296296295, | |
| "grad_norm": 52.15943145751953, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 12.7827, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 49.29362106323242, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 13.4053, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.859259259259259, | |
| "grad_norm": 48.308074951171875, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 13.3717, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.8740740740740742, | |
| "grad_norm": 48.63944625854492, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 13.6078, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 54.993064880371094, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 13.5741, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.9037037037037035, | |
| "grad_norm": 56.49517822265625, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 12.8867, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.9185185185185185, | |
| "grad_norm": 51.71754455566406, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 12.7084, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.9333333333333336, | |
| "grad_norm": 63.62958908081055, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 13.9539, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.948148148148148, | |
| "grad_norm": 57.946311950683594, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 12.6946, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 49.338043212890625, | |
| "learning_rate": 0.0, | |
| "loss": 13.4586, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "eval_loss": 3.4680771827697754, | |
| "eval_runtime": 1.304, | |
| "eval_samples_per_second": 87.423, | |
| "eval_steps_per_second": 22.239, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4874493864378368e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |